# ST2304 Statistisk modellering for biologer/bioteknologer (våren 2022)

This page contains the timetable and links to all of the material (modules, exercises etc.). There will only be small changes from 2021: this is set up with 2021 material, and we will update it week by week.

Course material (lecture notes etc.) from last year are here. The material for this year will be almost the same

For å få full tilgang til emnet, så må du

- ha en aktiv NTNU-brukerkonto
- ha betalt semesteravgift for våren 2022
- være undervisningsmeldt i emnet (sjekk på StudentWeb)

Kontakt faglærer Bob O'Hara hvis du fremdeles ikke har tilgang.

To hand in exercises – go to Blackboard. Emnet går i Blackboard Vår 2022.

There is a discussion board on Blackboard, you can find it in the Learning Materials. You should be able to post anonymously (if not, tell me and I'll try to sort it out)

## Exam Preparation

We will continue the virtual sessions on Tuesday (16-18) and in person on Wednesday (12-14) up to the exam date. If you want to cover a topic in more detail, just ask (on email or Blackboard).

Tuesday 16.15-18.00: online (https://ntnu.zoom.us/j/92050671785?pwd=OEx2R2M0L1RTdkxPQUt2OCs2Z3lvdz09: the same Zoom link as always) Wednesday 12:15-14:00: Room KJL2. This is in the same building as the Wednesday sessions, but on the ground floor: directly under the room we were in.

Exams and solutions from previous years are here

Exam FAQ:

- The exam is written in English, but can be answered in Norwegian. This isn't a language exam, so do what is comfortable. You can even change language between questions. And don't worry about "bad" English: if I can understand what you wrote, it's fine. - The exam will be like the ones in the last couple of years. - There will be no need to do any R programming in the exam.

### Timetable

All teaching will be online, at least for the first few weeks. We will re-assess as the advice changes.

(starts Tuesday 11th January)

- Wednesday 12:15 - 14:00 Zoom Link (if we meet in person, it will be in KJL5 Kjelhuset)

All materials will be found at this web link - just click on the correct folder for the lecture week https://www.math.ntnu.no/emner/ST2304/2022v/

### Recommended texts

None of these books is perfect for this course. If you want only one, Hector is probably best unless you are happy to read something a bit more technical (= has lots more equations), in which case Dunn & Smyth is good.

New Statistics with R - Hector: a useful practical book, it covers the material well, but not deeply.

The Analysis of Biological Data - Whitlock & Schluter - covers a lot of the more basic ideas in detail.

Generalized Linear Models With Examples in R - Dunn & Smyth This is mathematically a bit more advanced, but is fine if you are comfortable reading equations.

### Course Material

#### Week 1: Introduction (week beginning January 10th)

Videos These are from the first lecture, and lightly edited. The picture looks a bit off because I had to crop it:

These are pre-recorded:

#### Week 2: Maximum Likelihood (week beginning January 17th)

Starting this week you just go through the modules, so everything is pre-written and pre-recorded.

Videos (these are linked to in the module)

#### Week 3: Uncertainty in Maximum Likelihood Estimates (week beginning January 24th)

#### Week 4: The Normal Distribution (week beginning January 31st)

Note: there are some horrible-looking equation in the module. You won't have to know these for the exam. If you understand the approach we are taking, you don't have to follow the maths. At least you'll know it's not a black box.

- Exercise 2 (to be handed in by the end of Friday 11th February. Ignore what it says on the page)

#### Week 5: Regression (week beginning February 7th)

- Exercise 3 (the hand-in date is wrong: I will correct this soon)

To use the intro to markdown: download the file onto your computer (right-click and chose "save link as". Check that it saves with .Rmd at the end of the file name: that is how the computer know it is an R Markdown file). Then open it in RStudio (it might be enough to double-click on it). Then you have the markdown file in RStudio, so you can edit it. When you want to convert it to html, click the "Knit" button above the pane with the file in it. That should convert the Markdown document to html, and open a browser. If you click on the arrow next to the Knit button, you can get a menu so you can decide to knit the document to be a pdf or Word doc.

This course was written with Markdown, so now you know how it was done.

#### Week 6: Regression (week beginning February 14th)

#### Week 7: Multiple Regression (week beginning 21st February)

#### Week 8: Categorical Covariates (week beginning 28th February)

#### Week 9: More Categorical Covariates (week beginning 7th March)

#### Week 10: Model Selection (week beginning 14th March)

This week we split the module into 3 parts.

#### Week 11: Examples of "full" analyses (week beginning 21st March)

This week's material is a recap of what you have done, and putting it in the context of a proper analysis.

There is no separate exercise this week. You can consider these problems as exercises, if you want: if you submit them as exercises they will count towards the total exercises.

#### Week 12: Generalised Linear Models (week beginning 28st March)

(the week numbering might be out on some of the files: this was Week 11 in previous years)

Week 13: Binomial Generalised Linear Models (week beginning 4th April)

- Binomial Exercise (note: you should be getting an extra week to finish this exercise, because of Easter)

Week 14: Poisson Generalised Linear Models (week beginning 18th April)

### 2021 Material

CHANGES WILL BE MINOR, SO THERE IS NO HARM IN LOOKING AT THESE NOW

Week 14: Poisson Generalised Linear Models (week beginning 4th April)

Easter: Week beginning 11th April

Week 15 answers

- Cow Problem, alternative Answers (this covers some ideas we will not covered in the course, but which are important in many real problems)

Week 16 and later: Revision for the course. (week beginning 25th April)